A new diagnostic test for tuberculosis (TB) that is inexpensive, easy to perform and is both sensitive and specific is a high priority. Since the diagnosis of TB requires integrating laboratory and clinical performance standards, the evaluation of new diagnostic tests for TB is a challenge. Further, since there are many presentations of TB, a new diagnostic test may work better in some presentations of TB than in others. Physician evaluations can help to identify clusters, or subsets, of patients where a particular test can be most useful. The Clinical Suspicion of TB (CSTB) instrument provides a summary estimate of the pre-treatment likelihood of TB based on recognized epidemiologic, clinical, radiographic, and bacteriologic criteria for establishing a presumptive diagnosis of pulmonary TB (PTB). The CSTB has been used by this group in preliminary studies to determine the pre-treatment prevalence of TB. The first objective of this study is to characterize the physician's estimation of prior probability of PTB in different settings using the CSTB form as a technology assessment tool. The second objective is to apply this tool in evaluating the clinical performance of serodiagnosis as an aid to empiric treatment decisions. Serodiagnostic tests for TB provide an excellent test case for the CSTB in that many investigators who have evaluated these tests have found good operating characteristics, but not quite good enough for routine use. In this prospective, randomized, controlled trial, 3,540 TB suspects (600 expected to have PTB) will be enrolled at seven internationally representative clinical sites. The CSTB will be characterized at each site and the clinical performance of two new test formats for serodiagnosis of TB will be evaluated. Patients will be selected so as to study CSTB and test performance under both diagnostic and screening indications for PTB. Among patients randomized to receive on-site serodiagnosis or standard work-up without serodiagnosis, CSTB scores and the accuracy of empiric treatment decisions will be compared. Complementary laboratory studies will evaluate antigen recognition profiles in different populations. The specific aims are: 1) to evaluate the reliability and validity of the CSTB as a tool for estimating physicians' pre-treatment likelihood of disease, 2) to evaluate sensitivity, specificity, and predictive values of serodiagnosis under different testing indications, 3) to characterize operating characteristics of serodiagnosis using the CSTB, 4) to evaluate the impact of serodiagnosis on physicians' estimates of disease likelihood and empiric treatment decisions, and 5) to develop a decision model for estimating the clinical utility of serodiagnostic tests. Serum collected in this trial will be used to create a serum bank for testing of newer, as yet to be developed, serodiagnostic tests. This study will: 1) validate and streamline the CSTB as a simple tool to bring the richness of the clinical exam to the evaluation and use of laboratory tests for TB, 2) provide insights into how two serodiagnostic tests can be used to assist in rapid diagnosis of TB, and 3) establish a serum bank for screening of future serologic tests for TB. It will help define the role of serologic testing in the diagnosis of PTB. [unreadable] [unreadable]